Feb., 1931
POLAROGRAPHY OF THE ALKALINE DECOMPOSITION OF STREPTOMYCIN
585
ions.' It is therefore reasonable t o employ other after a little practice by both methods. Similar metals forming distinctive precipitates with 8- titrations using zinc as a fluorescent indicator were quinolinol as indicators in the titration of copper also tried but did not give good results. with a standard solution of 8-quinolinol in a manner Acknowledgment.-This work was supported in analogous to the Mohr method for determining part by a grant from the Cottrell Research Corhalide. The procedure was tried using both iron poration. and aluminum as the indicating ion; with ferric Summary ion the end-point is observed as the appearance of a black color; with aluminum a fluorescence 1. A procedure for the estimation of the relative marks the end-point. Results of fair accuracy stabilities of the chelates formed by several sub(Table 1) on prepared samples could be obtained stituted 8-quinolinols with copper has been described, and a conductometric titration method for TABLE I determining copper with 8-quinolinol has been MOHR-TYPE TITRATIONS OF COPPERWITH 8-QUINOLINOL developed. Indicator M g . Cu taken Mg. Cu found No. of trials 2. A volumetric method of analysis for copper Fe + 3 6.58 6.59 * 0.05 15 based on the relative stabilities of the various metal ~ 1 + 3 6.66 6.68.t .02 20 chelates of 8-quinolinol has been shown to be Zn +2 6.63 6.8 .3 10 satisfactory. f
[CONTRIBUTION FROM
THE
RECEIVED JUNE 26, 1950
LOUISVILLE, KENTUCKY
(1) Mellor and Maley, Nalure, 159, 310 (1947).
FRICK CHEMISTRY LABORATORY, PRINCETON UNIVERSITY ]
A Polarographic Investigation of the Alkaline Decomposition of Streptomycin BY CLARKE. BRICKERAND W. AUBREYVAIL A rapid polarographic method for assaying streptomycin has been reported by Levy, Schwed and Sackett.' These authors suggest a 3% solution of tetramethyl ammonium hydroxide for the supporting electrolyte for the polarographic determination of this compound. They also state that the wave height is a function of the PH of the medium and that the wave height is not constant until a PH of 13.6 is reached in the supporting solution. No reference was made to the polarographic behavior of the similar compound, mannosidostreptomycin. In this investigation a more thorough study has been made of the effect of pH on the diffusion current obtained from streptomycin (to be called streptomycin A hereafter) and from mannosidostreptomycin (to be called streptomycin B). Since streptomycin A or B which has been decomposed by heating an alkaline solution of these compounds was found to be non-reducible a t the dropping mercury electrode, it was decided to study the rate of decomposition of these compounds polarographically. From these decomposition studies, it has been possible to calculate the energy of activation for these reactions and by comparing these values with that obtained from a similar study with ahydroxyisobutyraldehyde, the rate determining step in the decomposition of the streptomycins has been deduced. Experimental Apparatus and Materials.-A Leeds and Northrup Model E Electrochemograph was employed for recording all polarograms. A polarographic cell assembly similar t o that described by Furman, et aZ.,2was used throughout this study. This assembly was modified according t o the apparatus described by Linganes so that the mass of mercury flowing (1) G. B. Levy, P. Schwed and J. W. Sackett, THISJOURNAL, 68, 528 (1946). (2) N. H. Furman, C. E. Bricker and E. B. Whitesell, Anal. Chcm.,
14, 333 (1942). (3) J. J. Lingme, i b d , 16, 329 (1944).
per second could be measured easily a t any applied potential. In addition a constant temperature jacket was used to maintain the temperature of the solution being polarographed. With this jacket, the temperature of the solutio: was maintained t o 1 0 . 2 ' at as high a temperature as 75 and to *0.05' at 25". Two capillaries were used in this study. All of the kinetic studies were made with a capillary that delivered 2.39 mg. of mercury per second. Those polarograms which were taken t o illustrate the effect of PH on the reduction of streptomycin were recorded with a capillary that delivered 2.00 mg. of mercury per second. Whenever it was necessary to calculate ID,the actual values of m and t were measured a t the potential which was used for measuring id. A saturated calomel electrod: was used as the anode for For the higher temperaall polarograms recorded a t 25 tures, a quiet pool of mercury was used. The potential of this anode was measured against a saturated calomel electrode with a Leeds and Northrup student type potentiometer. All pH measurements were made with a Leeds and Northrup research model PH meter using a KO.1199-30 glass electrode. Tank nitrogen for removing dissolved oxygen from all solutions prior to the polarographic analysis was purified by passing it through vanadous sulfate solution as described by Meites and M e i t e ~ . All ~ solutions were deaerated for 10 minutes with a slow stream of nitrogen prior to recording the polarograms. In the kinetic studies, the buffer solutions alone were deaerated ten minutes and then two minutes more after the solution of the reducible material was added. The phosphate buffer solutions were prepared by mixing NaHzP04 and Na2HP04 or by mixing NazHPOa and NaOH so that the total available phosphate concentration was 0.25 iM in all cases. The borate buffers were prepared from boric acid and sodium hydroxide with a resulting available borkte concentration of 0.1 M. A 1.0% gelatin solution containing a trace of mercuric iodide was used as a maximum suppressor throughout this study. This solution was perfectly stable and showed no evidence of mold growth over a period of several months. The streptomycin A and streptomycin B were specially purified and were obtained from the Heyden Chemical Corporation. Experiments with Streptomycin A and Streptomycin B.A 1.53 millimolar solution of streptomycin A trihydrochloride and a 1.54 millimolar solution of streptomycin B sulfate
.
(4)
L. Meites and T. Meites, ibid.. 10, 984 (1948).
586
1'01. 7 3
CLARKE. BRICKERAND IY.AUBREYVAIL
were prepared. Polarograms were recorded from solutions first wave is due to thereduction of the essentially free (probwhich contained 2.00 ml. of the streptomycin solution, 0.10 ably hydrated in solution as an aldehydrol) aldehydic group ml, of the gelatin solution and 10.0 ml. of the appropriate and the second wave is caused by reduction of a partially buffer or 10.0 ml. of a sodium hydroxide solution. A hound aldehydic group. (In streptomycin trihydrochloride saturated calomel electrode was used as the reference anode solutions a t pH's less than 12.5, the carbonyl group appears in all of these runs. The pertinent data from the polaro- to he bound by either or both the protonated guaniditlo grams recorded from streptomycin A are tabulated in Table groups or the protonated secondary amine.) This latter I. Streptomycin B gave the same results as streptomycin A reduction is probably not quantitative and therefore the I n values in Table I for the total of the two waves are not with the exception that a $H of 8 and higher, the I n values for the streptomycin B were slightly lower than those given constant. In other words, the two ID values for streptoin $H depend on the relative quantities of in Table I. entially free and partially bound aldehydic groups that TABLE I DATAFROM THE POLAROGRAPHIC REDUCTIOX OF STREPTO- Kinetic Study with Streptomycin A and Streptomycin B at Various Temperatures and pH.-A 10.00-1111.portion of a MYCIN AT VARIOUS PH buffer solution arid 0.10 ml. of 1% gelatin were placed it1 a I D (id/Cm*/a t ' / s ) B1lZ TIS. S.C.E. clean, dry, thermostated polarographic cell and deaerated for for for ten minutes. Then 2.00 nil. of approximately 1.5 1st wave 1 s t wave Total waves fiH inillimolar streptomycin A or B solution was added and the -1.68 14.0 1.56 1.56 deaeration was continued for two minutes more. As soot1 -1.56 2.28 2.28 13.0 as the streptomycin solution was added, a stop watch was started. At various time intervals after the additional t w o 12.4 -1.536 1.92 2.05 minutes of deaeration, polarograms were recorded from this -1.476 11.2 1.35 1.68 solution. -1.40 9.93 1.01 1.49 The Leeds and Sorthrup Model E Electrochernograph -1.38 0.98 1.48 9.40 was especially suitable for this study since the instrument is synchronized so that it requires 30.0 seconds to trans-1.32 ,79 1.16 8.68 verse 0.100 volt, Thus by starting the polarogram a t a .F1 1.00 -1.31 8.43 definite time, it tvas very simple to calculate the elapsed -1.28 8.00 .55 0.79 time between the addition of the streptomycin solution to 7.73 -1.26s .39 .58 the buffer medium and the time when i d was reached. I n order to record a sufficient range of a polarogram to facilitate --1*246 .29 .41 7.38 the measurement of the i d values and in order to rewind the -1.22; 7.13 .2x .40 chart paper and reset the polarizing unit and zero adjust.26 .34 -1.16 5.92 ment on the recorder, polarograms could not be recorded 4.20 .15 .l5 -1.04 more frequently than every five minutes. For very fast rates of decomposition, it is possible t o stop the polarizer a t The El/, values from Table I show a linear relationship a potential which produces the diffusion current and then with pH and can be represented by the equation, El/, = follow the change in i d continuously from the recorder. -0.806 - 0.0595 pH. When the IDvalues for the first The polarograms from the decomposition study on strepwave are plotted against fiH, a graph is obtained which is tomycin A in a phosphate buffer of pH 10.65 and a t a temvery similar t o the actual titration curve of streptomycin perature of 65" are shown in Fig. 1. A similar series of trihydrochloride with sodium hydroxide. I t was also ob- polarograms were recorded from streptomycin A a t 35, 45, served that the second polarographic wave from streptomy- 55 and 75' with the same buffer. In addition, decomposicin appeared t o coalesce with the first wave when the po- tion studies were made a t most of these same temperatures larograms were recorded from solutions a t higher tempera- using 1 N sodium hydroxide and other phosphate buffer:; tures. These facts can be interpreted on the basis that the having pH's of 10.3, 11.2 and 11.65, respectively.
0.q-
Fig, 1.-Polarograms of streptomycin A in phosphate buffer #H 10.6 at 65 O; the i d for polarogram 1 was recorded 10.0 minutes after the addition of the streptomycin A to the phosphate buffer solution; the time interval between polarograms was 5.0 minutes.
0
I
,
,
,
I
60 90 120 150 Time in minutes. Fig. 2.-Results of decomposition studies of streptomycin A in phosphate buffer of pH 10.6 plotted as logarithm of wave height versus time.
30
Feb., 1951
POLAROGRAPHY OF THE ALKALINE DECOMPOSITION OF STREPTOMYCIN
After it was proved that i d at a given PH and temperature was proportional t o the concentration of streptomycin, it seemed valid to use the wave heights directly t o interpret the decomposition data. From each series of polarograms, the logarithm of the wave height was plotted against the time elapsed between the addition of the streptomycin solution and the time a t which i d was recorded. The plot of the logarithm of the wave height against time for all of the decomposition studies made on streptomycin A with the phosphate buffer of PH 10.65 are shown in Fig. 2. Since this figure shows that the logarithm of the concentration of streptomycin varies linearly with time, the decomposition of streptomycin A must follow a f i s t order reaction. The slopes of these lines as well as those for all of the decomposition studies made on streptomycin A were calculated and are tabulated in Table I1 as the velocity constant, k, of the reaction (measured slope X 2.303 = k). Decomposition studies were also made on streptomycin B in two of the phosphate buffers and in 1 N sodium hydroxide. The velocity constants for all of these experiments are also listed in Table 11. It is apparent from the velocity constants where the same conditions were used for both streptomycins that both compounds decompose a t essentially the same rate.
phosphate media. Furthermore, the velocity constants for streptomycin B appear t o be consistently lower (from 2 t o 15%) than those for streptomycin A under the same conditions. Such a difference was not observed in phosphate buffers. Although the differences in velocity constants observed for streptomycin A and B in borate buffers are probably larger than experimental error, they are not sufficiently different to utilize for analyzing mixtures of these two compounds in the presence of each other. In order to calculate the experimental energy of activation (EeXP.)for the decomposition of the streptomycins, the
4.0
3.0 i
TABLE I1 VELOCITY CONSTANTS FOR THE DECOMPOSITION OF STREPTOMedium
1 N hTaOH
Phosphate buffer PH 11.65
MYCINS Velocity constant, k m h - 1 for Temp., OC. Streptomycin A Streptomycin B
25.0 30.0 35.0 40.0 35.0 40.0 45.0
50.0 Phosphate buffer PH 11.20
Phosphate buffer PH 10.65
Phosphate buffer PH 10.30
Borate buffer PH 10.70
Borate buffer PH 9.60
55.0 35.0 45.0 55.0 65.0 35.0 45.0 50.0 55.0 65.0 75.0 35.0 45.0 55.0 65.0 75.0 45.0 55.0 59.0 62.0 55.0 59.0 65.0
0.0062 .0131 ,0378 ,0627 .0026 ,0054 ,0101 .0216 .0437 ,0010 .0044 .0198 .0695 ,00026 .00124
.... .00629 ,0363 .0723 .00014 ,00069 .00262
.0138 .0485 .00107 .00442 ,00781 ,0134 ,00031 ,00058 .00130
0.0068 .0136 ,0355 ,0635 .0027 ,0053 ,0111
.... .0461
.... .... ....
....
.00023 .00128 .00445 .00668 ,0327
587
sM I
2.0
1.0
t I
I
I
I
I
I
i
I
2.96
3.15 3.35 1/T x 108. Fig. 3.-Logarithm of rate constant for decomposition of streptomycin A in various media uersus the reciprocal of the absolute temperature; squares, results at PH 10.3; open circles, PH 10.65; closed circles, PH 11.2; triangles, pH 11.65; crosses, PH 14.0.
.... .... .... ....
.... .... .00099 ,00374 .00760 . ,0127 ,00030 .00056 ,00127
One of the objectives of this study was to determine if the decomposition was only PH dependent on whether it was affected by the buffer anion. Since streptomycin B contains more hydroxyl groups than streptomycin A and since polyhydroxy compounds form complexes with borate ions, a borate buffer should show more effect than any other buffer solution. For this reason the decomposition of streptomycin A and B was studied over a temperature range using borate buffers of PH 9.6 and 10.7. The velocity constants from these studies are also presented in Table 11. These data when compared with those from phosphate buffers at the same pH and temperature indicate that the streptomycin decomposes at a slower rate in borate media than in
I
I
I
2.95
I/T
I
x
1
3.15
I
I
3.35
103.
Fig. 4.-Logwithm of rate constant for streptomycin A and B in various media versus the retiprocal of the absolute temperature; streptomycin A, open symbols; streptomycin B, closed symbols; squares, p H 10.65; circles, $JH 11.09) triangles, pH 14.0.
588
CL.%RK E. RRICKER AND IT.AUBREY\'AIL
logarithms of the velocity constants were plotted against the reciprocal of the absolute temperature. These data for streptomycin A in phosphate buffers and 1 N sodium hydroxide are shown in Fig. 3. This figure shows that the slopes of the straight lines are practically identical which indicates that the energy of activation is a constant value and independent of PH. Furthermore, since the energy of activation remains so nearly constant over the comparatively wide PH range, it is likely that the same mechanism for the decomposition of streptomycin occurs in all cases. That is, the reaction is probably not a pseudo first order reaction involving hydroxyl ions as well as streptomycin. The average value of E,,,. which was calculated from thc slopes of the lines in Fig. 8 was 30.2 * 0.7 kcal. To illustrate the great sitnilarity of the decomposition of streptomycin A and B, the logarithm of the velocity constants for the decomposition of both of these compounds where they were studied in the same media are plotted against the reciprocal of the absolute temperature in Fig. 4. This figure shows very clearly that these two compounds decompose a t the same rate and that they both apparently require the same energy of activation. The velocity constants for the decomposition of streptomycin B in borate buffers were found to be somewhat lower than those for streptomycin A. However, when the logarithms of these velocity constants for both streptomycins were plotted against the reciprocal of the absolute temperature, a pair of parallel straight lines was found for each borate buffer. These lines also had the same slope as those shown i n Figs. 3 and 1. It is necessary to conclude that, although the two streptomycins probably form coniplexes of different stability with borate ions which affect their rate of reaction, the decomposition which actually takes place requires the same energy of activation and must proceed by the same mechanism. The various thermodynatnic quantities, AH*, AF* and AS*,were calculated for the decomposition of the two streptomycins in the various media and their average values are given in Table 111.
TABLE 111 THERMODYNAMIC QUANTITIES FROM
1 N NaOH Phosphate buffer PH 11.65 11.20 10.66 10.30 Borate buffer pH 10.70 9.80
29.2 29.2 22.8 22.8 20.8 21.0 27.8 27.8 20.7 20.7 21.8 22.2 28.7 , . 20.2 . . 25.9 . . 29.1 29.1 20.3 20.3 26.6 25.9 .. 20.3 . . 29.4 . . 29.8 29.4 29.4 20.5 20.6 26.9 26.7 30.9 31.0 20.5 20.5 31.2 31.3
Experiments with a-Hydroxyisobutyraldehyde.-Schenck and Spielman6 observed that maltol was formed from streptomycin when this compound was treated with aqueous alkali under relatively mild conditions. Kuehl, et d . , O stated that maltol must be derived from the streptose moiety of the streptomycin. They also proved that a group must be attached glycosidically a t carbon-1 in streptose and that a free or potentially free streptose aldehydic group is necessary in order t o get maltol. Lemieux and Wolfrom' have suggested a possible series of reactions for the alkaline degradation of streptomycin to maltol. The first reaction in this degradation is similar to the rearrangement of ahydroxyisobutyraldehyde t o form acetoin which was studied by Danilov and Venus-Danilova.8 Since it seemed likely that the a-hydroxyisobutyraldehyde would be reducible ( 5 ) J. R. Schenck a n d M. A. Spielman, TRISJOURNAL, 67, 2276 (1945). (6) F.A. Kuehl, Jr., E.H. Flynn, N. G. Brink and IC. Folkers, i b i d . ,
68, 2879 (1946). (7) "Advances in Carbohgdrrlte Chemistry," edited by W. W. Pigmaii and M. L. Wolkrom, Vd. 8. Acadetnk Press. h c . , New York, 1948. ehapter hbtitied "The Chemistry bt' Strdptomycid" by R . U . Lemieux nhd I.. %olF$drn. (SI bairitoy aird E. ~ e u i i s - ~ e o i i o vBW., ;~, 64 ( l ~ 2 4 ) .
M.
a t the dropping mercury electrode and that the acetoin would not undergo reduction a t the same potential, it was decided to investigate the kinetics of this rearrangement so a comparison could be made with the data from the streptomycins. The a-hydroxyisobutyraldehyde was made from a-bromoisobutyraldehyde by hydrolyzing the latter compound and neutralizing the resulting solution with barium c a r b ~ n a t e . ~ Because of the difficulty of separating and preserving the pure a-hydroxyisobutyraldehyde, the hydrolyzed solution of a-bromoisobutyraldehyde was filtered and used without further purification. All kinetic measurements were made i n the same manner as described previously. From the rate of rearrangement of a-hydroxyisobutryaldehyde, which was measured a t several temperatures in phosphate buffers of pH 10.3, 10.65 and 11.2, respectively, the velocity constants and various thermodynamic quarititie.; were calculated. These values are listed in Table IV.
1)ATA FROM
TABLE IV KINKTIC STUDIES WITH a-HYDROXYISOBUTYRALDEHYDE
Phosphate buffers
P
10.30
10.RR
11.20
DECOMPOSITION
STUDIES ON THE T W O STREPTOMYCINS A refers to streptomycin; B to mannosidostreptomycin. AH*, kcal. A F * , kcal. AS*, e.u. Medium A B A B A B
Vol. 72
Eexp., kcal.
AH*, kcal.
AF*,
AS*,
kcal.
e.u.
45.0 0.00375 31.1 50.0 .00972 3 1 . 1 55.0 .0180 31.1 .0336 31.1 60.0 65.0 .0778 31.1 47.0 .0119 30.8 50.0 .0174 30.8 55.0 ,0368 30.8 35.0 ,0195 31.1 ,0245 31.1 37.0 45.0 .0974 31.1
30.5 30.5 30.5 30.4 30.4 30.2 30.2 30.2 30.5 30.5 30.5
19.4 19.1 19.1 18.3 18.7 19.3 19.3 19.1 19.7 19.5 18.7
34.6 35.0 34.8 35.7 34.7 33.7 33.4 33.7 35.1 35.3 37.0
Temp., "C.
k, min.-1
Discussion
The average AH* values for the decomposition of the streptomycins (Table 111) are only 1.1 kcal. different from the average A€€*value obtained from the rearrangement of a-hydroxyisobutyraldehyde (Table IV). This would suggest that the rate determining step in the decomposition of streptomycin must be similar to the rearrangement observed in the isobutyraldehyde and therefore must involve a change of the aldehyde group in the streptomycin. This indicates that the first step in the mechanism proposed by Lemieux and Wolfrom? for the production of maltol from streptomycin is probably the rate determining step in the alkaline decomposition of this compound. It has been shown that the buffer anion exerts some effect on the rate of decomposition of the different streptomycins but this effect is very small in proportion to the PH effect. The decomposition of the two streptomycins and the rearrangement of a-hydroxyisobutyraldehyde show essentially the same PH dependence. The logarithm of the velocity constants for all three of these compounds changes about one unit for every unit change in PH between 9.6 and 11.6. The fact that these compounds show the same pH dependence provides additional evidence that the same general reaction must be responsible for the rate determining step in both the decomposition of streptomycin and rearrangement of a-hydroxyisobutyraldehyde. The velocity constant for the rearrangement of ahydroxyisobutyraldehyde was observed to be considerably higher than that for either of the strepto(D) A. Franke, Monnfsk,, 91, 2.71 (lHfl0).
Feb. , 1951
STRUCTURE OF
ERYSOPINE, ERYSODINE AND ERYSOVINE
mycins under similar conditions. This is not surprising if it is assumed that an essentially free carbonyl group is necessary for the rearrangement. The aldehyde group in the streptomycin molecule at a PH of 12.5 or less is apparently partially bound. Thus, in solutions of these PH's, every streptomycin molecule does not produce an essentially free aldehyde group which is capable of rearrangement. Consequently, the rate of decomposition of streptomycin should be slower than the rearrangement' of a-hydroxyisobutyraldehyde where all the aldehyde groups are essentially free. Acknowledgment.-The authors wish to thank the Heyden Chemical Corporation for supplying all of the streptomycin that was used in this investigation. Summary 1. A study has been made of the polarographic reduction of streptomycin and mannosidostreptomycin over the PH range of 4.2 to 14. 2. The rate of decomposition of streptomycin has been followed polarographically in various alkaline phosphate and borate buffers and in 1N sodium hydroxide.
[CONTRIBUTION FROM THE
589
3. In the PH range 9.6 to 14, the rate of decomposition of both streptomycins has been shown to be identical in phosphate buffers and to follow a Grst order reaction. The rate of decomposition for these compounds differs slightly in borate buffers. The energy of activation, free energy of activation and entropy of activation were calculated for this reaction. 4. The rate of rearrangement of a-hydroxyisobutyraldehyde to form acetoin was measured and compared with the rate of decomposition of streptomycin. Both of these reactions are first order, require essentially the same energy of activation, and show the same pH dependence. It was concluded that the rate determining step in the alkaline decomposition of streptomycin is probably the rearrangement of the streptose moiety to form a sixmembered ring. 5. Because of the great similarity of the two streptomycins in their polarographic behavior and in their rates of decomposition, it is unlikely that a simple polarographic technique can be used to determine these compounds in the presence of each other. PRINCETON, NEWJERSEY
RECEIVED AUGUST7, 1950
RESEARCH LABORATORIES OF MERCK&. CO., INC.]
Erythrina Alkaloids. XVIII. Studies on the Structure of Erysopine, Erysodine, Erysovine and Erythraline BY KARLFOLKERS, FRANK KONIUSZY AND JOHN SHAVEL, JR.
It was established' that erysopine, erysodine and erysovine have the same four-nuclei ring system and that each has three oxygen atoms a t the same positions on the ring system. However, these three alkaloids differ in the number and position of 0-methyl groups. These deductions are based on analytical, hydrogenation, and methylation experiments, and provided a basis I, R = CH30- and >C=C< IV for the new Hofmann degradation and other 11, R = CH80- and 2>C=C< 111, R = CH& HO- and > C=C < reactions which are described herein. The structures2 of erythramine (I), erythraline Hz H2 (11) and erythratine (111) also provided a basis for further work on erysopine, erysodine and erysovine since all of these alkaloids seemed to have related structures. The conversion of laudanosoline into dehydrolaudanosoline chloride (IV) and then into 2,3,11,12-tetrahydroxydibenzoH2C H&tetrahydropyrrocoline (V) has been discovered VI V by both RobinsonS and Sch6pf4 and their coworkers. The properties of the pyrrocoline derivative (V) and the Hofmann and Emde degradations hydroxyl groups.' Each of these alkaloids has of it were promising precedents for these new two double bonds in addition to the benzenoid nucleus and yields tetrahydro derivatives.'." It Erythrina alkaloid studies. Erysodine and erysovine have the formula was desirable to aromatize the ring containing ClsH21NOs, and have two methoxyl groups and the two double bonds before attempting a Hofmann one hydroxyl group; erysopine has the formula degradation. This aromatization was accomC ~ ~ H ~ ~ Nand O S ,one methoxyl group and two plished by treatment of the alkaloid with hydrobromic acid which resulted in a decrease of CH40 (1) Koniuazy, Wiley and Folkers, Tms JOURNAL, 71, 875 (1949). in composition. The reaction of erysopine with (2) Folkers,Koniuszy and Shavel, ibid., 64, 2146 (1942). (3) Robinson and Sugasawa, J . Chem. SOC., 789 (1932). (4) Schllpf and Thierfelder, A n n . , 497, 22 (1932).
(6) Folkers and Koniuszy, THISJOURNAL, 81, 1677 (1940). (6) Folkers and Koniuszy, ibfd., )I, 1832 (1950).
